This application aims to provide an understanding of the chemical mechanisms of those types of enzymic and non-enzymic glycosyl transfer not yet understood. Two types of non-enzymic glycosyl transfer are involved: transfer by an internal return (SNi) mechanism, and reactions of derivatives with a carboxylic acid attached to the reaction center (not only sialic acids but also glycosides of 3-deoxy-D-manno-2-octulosonic acid (KDO). Transition state structures for these processes will be inferred from secondary deuterium (alpha, beta, and gamma), 180 and 13C kinetic isotope effects, measured by the isotopic quasi-racemate method using labelled glucosyl and N-acetyl neuraminyl derivatives as substrates, as well as from the usual techniques of physical organic chemistry. Glucosyl transfer to phosphate by internal return is the favored mechanism of glycogen phosphorylase, a key enzyme of mammalian primary metabolism. Sialyl residues on glycoproteins and glycolipids modulate many aspects of cell-cell and cell-protein interaction in mammals, and KDO residues are important in determining bacterial antigenicity and endotoxicity. Labelled NANA derivatives will be used to probe the mechanism of neuraminidase, to see if the carboxylate group of the substrate takes the place of the enzyme carboxylate group which acts as a nucleophile in other types of retaining glycosidase. Reaction with inversion of the anomeric configuration is the major area of ignorance in respect of enzymic glycosyl transfer. Transfer to phosphate (which is significantly reversible) will be probed by multiple kinetic isotope effects, measured by equilibrium perturbation, using bacterial disaccharide phosphorylases and E. coli purine nucleoside phosphorylase. Transfer to water (which is irreversible) will be investigated using Aspergillus glucoamylase, Trichoderma reesei cellobioside hydrolase II, and Bacillus pumilus beta-xylosidase. Those kinetic isotope effects amenable to direct measurements will be measured, and transition state analogues in which the (at present hypothetical) nucleophilic water molecule is mimicked by hydroxyl or fluoro groups will be synthesized. With mechanistic detail about inverting glycosidases to combine with what we know about retaining glycosidases, the question of why two stereochemical pathways are adopted when mutarotation makes them metabolically nearly equivalent will be addressed.